Why DevOps Jobs Will Keep Growing in 2027

DevOps engineering roles in the United States have a median total compensation of $150,000 a year. Cloud engineers are closed at $151,000 and SRRES make even more at $25,000 a year of median total compensation. In this video, I'm going to share why these people make so much money, why the demand for these people is not going to slow down, and it's only going to increase in next few years, and I'm going to give you the full road map that you can use to land a six-figure job like this. So for this video, I've prepared one of these famous Misha Vandenberg mind maps for you and we're going to be exploring this together. And the first element of the mind map is going to be familiar to you if you have watched any of my content in the past few years. But the first element is of course going to be Kubernetes. And already some of you as soon as you saw this logo, you went down through the comments and you're already starting to write an angry, hateful comment that Kubernetes is absolutely overkill. It's not needed. no one should learn Kubernetes because no companies use it, etc. But if there's going to be one thing you're going to be taking away from this video is that this is absolutely false and that if you want to have a DevOps, cloud engineering, S sur or infrastructure role in 2027 and if you want to do that without Kubernetes, you are not going to make it. It's that simple. You are not going to make it and you're not going to land that six-figure job. If you don't know what Kubernetes is, then of course we have to start there. But Kubernetes is basically the

operating system of modern application development and deployment. So modern applications always run containers and I'm going to explain more about that in a few minutes. But Kubernetes is the tool that large companies use to run these kinds of applications at scale. It has very quickly become the operating system of the cloud and AI. More on that later. But what you really have to know here is that Kubernetes is basically Linux under the hood. And that is going to be very important that you understand this in the next few sections of the road map. But if you learn Linux well, if you learn containers well, then Kubernetes is going to be very easy for you to pick up. However, the first hurdle here, the first thing that you need to get out of your mind is that Kubernetes is somehow overkill and that companies don't use it at all. And that's what we're going to get into now. Because in 2025 the CNCF the cloudnative computing foundation did a survey and this was released early 2026 and where their the companies that they surveyed mentioned that 82% of the container users deploy Kubernetes in production up from 66% in 2023. Okay. So Kubernetes is increasing in adoption amongst those people that were surveyed. Now the product manager of the Azure Kubernetes service. So this is a managed service that large cloud providers offer to their customers where they basically give you a Kubernetes cluster without any management. So it's a one-click

deployment solution for that that's called Azure Kubernetes service and the project manager of that product within Microsoft did a presentation at CubeCon. This is about a year ago I believe one and a half years ago where he stated that 95% of the Fortune 500 are using Azure Kubernetes service. So this is not just Kubernetes in general. This is even just Azure Kubernetes service. Okay, let me repeat that 95% of the Fortune 500 are using Azure Kubernetes Service. So, these are companies like the NBA, CocaCola, maybe you've heard of a small company like Coca-Cola, maybe a little company like Nvidia. Anyone heard of that? AT&T. What about Mercedes? Some kind of car brand. I believe it is small little company. 95% of the Fortune 500 are using Kubernetes. In other words, if you want to have an infrastructure job at these companies, you better be focusing on Kubernetes. Now what's even more interesting in that same survey is that they very clearly stated that Kubernetes is now the operating system of AI. So here the front page of that survey says the infrastructure of AI is future. And here there is they made a claim in that report saying that 66% of organizations use Kubernetes to host generative AI workloads. And here in the next section they even have the headline Kubernetes as the AI platform. And here it says the rise of Kubernetes as the de facto AI platform represents a

fundamental shift in how organizations approach machine learning operations. So the CNCF comes out saying that Kubernetes is the AI platform. Now remember this very carefully. Okay, let's repeat it. 66% of organizations use Kubernetes to host AI workloads and Kubernetes is the AI platform. Remember this carefully as we go into the next section. So if you go to LinkedIn right now and you enter Kubernetes and you search for worldwide on the jobs, you're going to see well over a 100,000 jobs available there. Now a large piece of this is coming from the application container market. I've been in this industry for several years now and I have seen this shift myself and increasingly it is being adopted more and more and any self-respecting modern company that runs software is running it in containers right now. We're going to get into this in a few minutes, but Google runs everything in containers and has been doing this for many, many years. And this of course has an effect on the market. So there's this company called Grandfue Research and they did a forecast on the application container market. And if you look at 2026, we are yeah maybe at 8 or 9 billion right now.

And in four years from now, this market is prospected to grow to $31 billion. $31 billion. That's a lot of money. Now, Kubernetes is a container orchestrator. So, if the application container market is going to grow, that means that the demand for Kubernetes is also going to grow. All of these these $31 billion are going to be running on Kubernetes because Kubernetes is the only sensible solution to run containers at scale. So, if the application container market grows, then Kubernetes grows. Now, the next piece that's driving all of this demand is of course AI. We've already seen that Kubernetes is the de facto operating system of AI. You don't need to be a philosophical genius to make the connection that if AI grows, then it needs to run somewhere. It needs to run on something. So if AI grows, then obviously the thing that AI is running on is also going to grow. The more AI models, the more AI users we get, the more Kubernetes clusters need to be operated. It it is really that simple. I am not some sort of prophet. This is not very sophisticated thinking. But people don't see this connection. Like whenever I talk to people and I explain this to them, it's like, oh wait, I never considered that. But if AI grows, then the thing that AI runs on also grows.

That's very important. So let's check out the growth of AI then because it's crazy. Like the AI competition is now limited by how quickly they can actually construct the data centers. So they don't have enough people to build these data centers. So that is already one of the bottlenecks. If we then look at how much this is going to grow over the next few years, McKenzie did research on this where they stated that by 2030, data centers are projected to require $6.7 trillion worldwide. So that's a nearly $7 trillion that is being invested in data centers specifically for AI. And again, these data centers are going to be filled to the brim with servers. And what's going to be running on those servers? Well, if you look at the data from the CNCF, those servers are going to be running Kubernetes. Now, Gardner also had a very interesting release on the 16th of September in 2026 where they forecast AI spending to grow with almost 50% in 2026 where convergence of AI infrastructure investment and AI embedding into software and services is driving AI spending to unprecedented levels. Worldwide spending on AI is forecast to a total $2.7 trillion in 2026. and where demand for AI infrastructure remains strong and inelastic to pressures from memory related pricing increases. In other

words, spending on AI is increasing with 50%. Like if you don't understand how significant that is, but if you're talking about these numbers, $2.7 trillion, like try to imagine $2.7 trillion, like our human brains are not even able to comprehend those kinds of numbers. Okay? But all I can say as a simple man is that when I see a market that's going to invest $2.7 trillion in one year into something, then obviously it's also going to lead to more jobs, more skilled people that need to operate these kinds of clusters. And finally, another interesting one from the 10th of September by IDC is that the worldwide server market revenue hits record 166 billion in the second quarter of 2026. And if you look carefully down in the bottom here, GPU accelerated servers generated almost 100 billion, which is a 30% yearover-year representing 52% of total market revenue. 52% of total market revenue on servers are GPU accelerated servers right now. Again, GPU accelerated servers are there to run AI. We've seen a Kubernetes is the operating system of AI. So these servers that you see being purchased here, they're all going to be running Kubernetes and there's going to be a surge of jobs for this stuff in the future. So by now you will have seen the evidence that this market is growing and

that if you want to have a DevOps or infrastructure related job, you must focus on Kubernetes. Don't come to me saying that Kubernetes is overkill or that it is not needed because it's simply false. And I know this because I'm a senior DevOps engineer who now specializes in coaching IT professionals into DevOps careers. And in the month of September alone, eight of my clients have landed jobs. That's around two jobs per week. We have Kevin, we have Erdinbot, we have Robert, we have Rodulfo. All of them have landed jobs using the same road map that I'm going to share with you in a second, which is constructed completely around learning Kubernetes and specializing in that. And that is the way into the DevOps careers these days. So we know Kubernetes is the way. Now we need to take a look at the rest of the road map. And as I mentioned before, if you want to have a Kubernetes job, it is absolutely essential that you have good Linux skills. You need to know Linux on a fairly deep level. You need to be able to work with it dayto-day and you need to understand it so that you are proficient on the command line. That is the most important thing because everything in modern DevOps is based on Linux. CI/CD pipelines, containers, Kubernetes, it's all based on Linux. And the main thing is that if you learn Linux well, then your life is going to be so much easier because when you are at an production outage on Friday afternoon, if things go wrong, then you're going to find yourself on a Linux

command line, logging into a server, debugging something right there. Like that is where the real work happens these days. There's going to be a failing container. You need to be going onto the Kubernetes cluster, checking out the storage. Where is this persistent volume? Why is it not being provisioned, etc. uh and all of that happens from the command line. In order to get good at the command line, you need to understand Linux well. You need to understand the file system, permissions, systemd, package managers, storage, all of this is extremely important to have good skills in. So that is where you start. Now you also go deep into bash scripting because CI/CD pipelines, which we're going to be getting into in a second, CI/CD pipelines are basically just glorified script runners. Okay? So if you understand bash scripting then you also automatically understand CI/CD pipelines to a very large degree. And when you have built up a decent set of Linux skills then the next step is to go to containers. Containers again like I said earlier everything at Google runs in containers. This is publicly available information. You can look it up for yourself. So having skills with Docker, knowing how to create a Docker file, container security, running multiple containers, Docker Compose, those are very important. Docker networking as well. Understand that container networking is just Linux foundations under the hood. Docker is just a wrapper around simple kernel features like croups, namespaces, network namespaces, etc. You have to understand that kind of level. And I say Docker here, but we're

also seeing the adoption of other container runtimes as well in the industry. Podman is used more and more and more. It's good for you to understand the difference between Docker and Potman, for example. If you cannot explain that then you don't know enough about containers in my opinion. But if you look at the docker website then it's going to be very clear that of course the industry is using this a lot. You see Lego, notion, Yamaha, Netflix, PayPal, Salando and Visa. They're all running containers and they're all running them at scale. So you can almost be sure that all of these companies are also running Kubernetes at this point. Now, so far, we've only been focusing on technology, but technology is not the only answer. And I'm telling you, I've been doing this for several years full-time now. I know how to get people into these careers. And from that experience, I've developed the job landing equation, which states that a job offer is the result of tech, reach, and proof. So, before I explain how the equation actually works, let me just show you what each of these mean. So tech obviously is what you've seen so far and also what every other single DevOps road map on YouTube is going to tell you. It's always just going to be a list of technology to learn, but it's not the full equation. That's not how you're going to land the jobs because you also need something called reach.

You can have all of the tech skills in the world, but if nobody knows you exist, you are never going to land the job. No one is going to come knocking on your door saying, "Hey, would you like to have a job?" Because nobody knows you exist. Now, most people only know one method of reach, which is to send job applications. But unfortunately, this is completely broken in the modern world. The strategy that most people use is to go on LinkedIn, click the easy apply button, and then you see over a 100 applicants, and that's what you're competing with. You're competing with hundreds of people who click easy apply, and there's no way for you to stand out there. So, in other words, you need new and other ways of reach. You need better ways of reach. And the final element here is proof because modern hiring has changed in the sense that they don't really care about what you had done 15 years ago. What they really care about is they want to see proof that you can do the job that they need you to do. It sounds very obvious, but it it has completely changed. I mean, Kubernetes is over 10 years old now, but computer science students are not learning about Kubernetes. There's no Kubernetes degree in any university. you have certifications but the only way that you can actually learn Kubernetes is to build stuff with Kubernetes. The only way that you can do it is to actually do it and use that as proof. So this is also reflected in the 2025 state of tactile report by the Linux Foundation where they were surveying hiring

managers and when they were asked how important are the following factors they mentioned that relevant Hansen experience is 95% importance portfolio and examples of previous IT accomplishments 85% where formal UN university degree is only at 65%. In other words, proof is more important than a degree or stuff that you have done 10 years ago. What you really need is proof that you can do the work. Now, going back to the equation then is you will see that this is a multiplication. So, someone who has extremely strong tech skills, he has been reading all of these books and he he has all of the theory in his mind, but he has no reach and he has no proof. He's only going to score a nine on the job equation. But the Chad down here who has only a six in tech skills, but he has really good reach and he has tons of proof of the skills that he has. Well, he is scoring super high on the equation. He's it's very likely that he's going to land a job. In other words, all of these elements need to be there. Tech, reach, and proof. Now, this is exactly what I built the CubeCraft DevOps road map around. There's a link in the description. And if you follow that, there's going to be a little survey that you fill out and then it's going to tell you exactly where you are stuck, whether you need more tech, whether you need more reach, or if you need more proof.

It's going to tell you where you are, what it feels like, and most importantly, how to get out of it, how to get into the next stage of your DevOps career. So, check it out. It's completely for free. There's the link in the description. When you fill it out, you're going to get a 45page document explaining everything to you in your tailored situation. Check it out. It's completely for free. The link is in the description below. So, how to get better reach and proof is explained in that document. But for now, we're going to go back to tech and focus on the tech road map here where the next section is going to be CICD. This stands for continuous integration, continuous deployment. And very briefly explained, CI basically means testing of code and CD means deployment of code. Now for the CI part, most companies these days are using GitHub and GitHub has their own solution for that called GitHub actions which is a tool in which you can build CI pipelines. So you can automatically test your code, you can build it, you can scan the container images, all that kind of stuff is typically done in GitHub actions. Some companies are using GitLab for this for example. But for learning purposes, I just recommend focusing on GitHub actions. And if you learn GitHub actions, then it's basically the same as all of the other stuff. So do that first. And then CD in again if you're focusing on Kubernetes that is a little bit of a different story. CD in modern application deployment basically means GitOps and the tools for to learn for that are either Flux or Argo CD. Now if

you are a real man and if you work on the command line then Flux is the tool that you want to use. Argo CD is also used a lot in companies. So you will have to learn it. But it's much more UI focused is probably also why it's used more. But Flux is obviously the superior tool here. And this is also what I recommend to my students is that they learn Flux first so that they learn the GitOps philosophy properly, learn how to work with it properly. And after that you can learn a less optimal implementation all of that by learning Argo CD which is going to be super easy after you've learned Flux. Now the next piece on the DevOps road map is going to be networking. It is very important that you have strong networking skills. I've been talking to senior networking engineers at Nvidia who mentioned to me that he is shocked at when people come in and they apply for DevOps engineering roles that they don't even have basic networking skills. So you definitely want to be able to explain cider ranges.

You want to be wellversed in container networking. So virtual Ethernet pairs, bridges, that kind of thing. Network name spaces and cloud networking. We're going to get into cloud in a second. But don't skimp on your networking knowledge. It's really important to have at least networking fundamentals down pattern. Next, there is programming. So yes, even in the age of AI, I still think it's important that you are going to learn fundamentals of programming if you want to land a DevOps job. Most of you who are watching this who and most of my clients are already in IT and I would say about 50% of my clients are already software engineers. So they can basically skip this section. If you are for example a support engineer, lots of my clients are support engineers or system administrators, then you don't necessarily have a lot of experience with writing code. Now, I think it's still important to learn at least the fundamentals of one language. I'm going to recommend that you learn Python and I'm going to recommend that you're going to learn how to write simple Python programs by hand in Vim first and later you can then use that knowledge to do all of the coding with AI. I mean, writing code has basically been solved by the large language models. I personally haven't written a line of code and I've built and deployed multiple applications myself during that time. All of them with all of the code written by AI. So, it's definitely solved, but I still think it's important

in in terms of concepts, right? You need to understand how functions work, how what variables are, etc. because this is going to get get back to you in infrastructure as code in cloud engineering and all of that stuff. So you don't have to become a DevOps engineer. There's no need to do lead code that kind of thing. But learn the fundamentals well and learn to write do some basic programming because it's going to help you to prompt these models. Like you need to learn how to set up a Python project with UV for example like do that by hand first and then you can prompt the AI models to do that later. But you need to go through it by hand first in my opinion. And the final piece on the DevOps road map then is going to be cloud. So I recommend that you pick one cloud and go deep. The biggest players are of course going to be Azure and AWS. Google Cloud is getting an a smaller increase of share from what I'm hearing, but you cannot go wrong with either Azure or AWS. One big mistake I see people make is that they are learning multiple because it will increase their chances of landing jobs.

I don't think that's true. I don't agree with that. It's best to pick one and go deep and you need to learn how to work with infrastructure as code. Now the tool for that is going to be Terraform and the advantage of Terraform is that you can use it on both AWS, Azure and many other places as well. However, Azure has a has its own domain specific language called Bicep and I really prefer bicep over Terraform any day. It's much more intuitive. It's much nicer to work with in my opinion. But Terraform is required and it's required that you know it. Now, if you want to go even deeper, then check out the CubeCraft DevOps road map. It's completely for free. You can just go to the link in the description. You fill out the little survey and then it's going to place you on the road map and it's going to give you advice specifically to your situation. It's going to tell you what your problems are and how to solve them. It's much deeper than what I went into in this video. So, check out the CubeCraft DevOps road map in the description below. It's completely for free. Thank you so much for watching and I'll see you in the next